Xtractor Depot put Ramped agents in Telegram and Gmail to clear inbox noise, surface decisions, and keep daily priorities moving.
A 10-year-old precision lab-equipment supplier serving demanding customers across regulated industries and advanced manufacturing. The founder did not need another chatbot. He needed a managed AI workflow that could triage Gmail, prepare daily briefs, and route approvals from his phone.
- Industry
- Precision lab equipment
- Founded
- 2016
- Stack
- Gmail · Workspace · Telegram
- Interface
- Telegram (mobile)
- Go-live
- 27 days from kickoff
The founder was the bottleneck — and the inbox was where it showed.
By 2026, Xtractor Depot was a fast-scaling precision lab-equipment supplier. Orders were up and the team was capable, but every important decision still routed through one person: Andrew, the founder.
The drag was not strategy. It was the manual work around it: email, follow-ups, status questions, and context scattered across a dozen tools.
- Inbox overload — Tens of thousands of messages and thousands unread, with real customer, finance, and legal items buried under newsletters and notifications.
- Context everywhere — Status, scheduling, and follow-ups spread across Gmail, Calendar, Drive, and the ERP, each with its own login.
- Nothing proactive — No layer surfaced what needed attention; Andrew had to go looking, usually after hours.
- Founder focus — The work only he could do, such as sales, product, and key relationships, kept getting crowded out by operational noise.
Andrew had tried AI tools before. Each worked for a weekend, then broke when an integration changed. None had the business context to keep helping without constant babysitting. That's when he started working with Ramped.
Day 1 to day 27. One workflow hub, then a team of agents.
Agent 1: Primary agent
Live on day 9Andrew's single point of command, reachable from Telegram on his phone. It holds his priorities, delegation rules, and communication style, then coordinates the specialist agents. Ask it anything from “what needs me today?” to “push Q3 revenue,” and it routes the work and reports back.
Agent 2: Inbox triage
Live on day 14Sorts, labels, and routes Gmail continuously, archiving repetitive noise like receipts, newsletters, and notifications while preserving every customer, revenue, finance, and legal thread. Nothing is deleted: high-signal items get surfaced for Andrew, the rest get routed or filed.
Agent 3: Accountability briefings
Live on day 20A morning briefing at 7:00 AM and an evening accountability check, plus a weekday business-objectives review across revenue, cash, marketing, and operations. Instead of Andrew remembering every open loop, the system brings the day's priorities and risks to him.
Agent 4: Specialist coordination
Live on day 27A roster of role-based agents for operations, finance, marketing, and revenue works from shared objective files, with knowledge refreshed across them every 30 minutes. The primary agent hands work to the right specialist and turns their results into one answer.
What needs me today?
Overnight triage is done. 38 sorted, 2 flagged for you: a customer escalation and a contract to sign. Nothing else is on fire before your 10am. Want me to hold 2 to 4pm for deep work?
Before and after, the first month.
The numbers are the visible part. The real change is that Andrew stopped starting his day in a 90,000-message inbox and started it with a briefing. The work only he can do stopped competing with everything else. Nothing was deleted; the noise was archived and routed, the signal surfaced.
“The biggest win was not that AI answered one email faster. It is that I can text it from my phone, clear the noise, surface the real decisions, coordinate the other agents, and see what actually needs me. It turned AI from a novelty into a real operating tool.”— Andrew, Founder, Xtractor Depot
The biggest win wasn't speed. It was getting his attention back.
Andrew expected an inbox cleaner. What he got was a managed workflow that changed how the business runs day to day:
Priorities and risks arrive proactively each morning, with an evening accountability check to close the loop.
Andrew asks from Telegram. The agent checks the connected tools and answers without making him open dashboards.
Operations, finance, marketing, and revenue agents share objective files and sync skills every 30 minutes.
Repetitive mail archived and routed automatically; customer, revenue, finance, and legal threads always preserved.
Next: deeper finance and revenue automation.
With the command layer in place, Xtractor Depot is extending what the specialist agents own:
- Revenue operations — Deal-stage hygiene, automated follow-up sequences, and pipeline risk surfaced to the primary agent before deals stall.
- Finance & cash — Recurring cash and margin reviews drafted from the books, with exceptions routed for a decision instead of sitting unseen.
Want results like this in your business?
Book a 30-minute AI workflow audit. We'll map your top three workflows and show the ROI math. Free, no commitment.